BindingSiteAugmentedDTA: Enabling A Next-Generation Pipeline for Interpretable Prediction Models in Drug-Repurposing (preprint)

While research into Drug-Target Interaction (DTI) prediction is fairly mature, generalizability and interpretability are not always addressed in the existing works in this field. In this paper, we propose a deep learning-based framework, called BindingSite-AugmentedDTA, which improves Drug-Target Affinity (DTA) predictions by reducing the search space of potential binding sites of the protein, thus making the binding affinity prediction more efficient and accurate. Our BindingSite-AugmentedDTA is highly generalizable as it can be integrated with any DL-based regression model, while it significantly improves their prediction performance. Also, unlike many existing models, our model is highly interpretable due to its architecture and self-attention mechanism, which can provide a deeper understanding of its underlying prediction mechanism by mapping attention weights back to protein binding sites. The computational results confirm that our framework can enhance the prediction performance of seven state-of-the-art DTA prediction algorithms in terms of 4 widely used evaluation metrics, including Concordance Index (CI), Mean Squared Error (MSE), modified squared correlation coefficient (rm2 ), and the Area Under the Precision Curve (AUPC). We also contribute to the two most commonly used DTA benchmark datasets, namely Kiba and Davis, by including additional information on 3D structure of all proteins contained in these two datasets. We manually extracted this information from Protein Data Bank (PDB) files of proteins available at https://www.uniprot.org/. Furthermore, we experimentally validate the practical potential of our proposed framework through in-lab experiments. We measure the binding interaction between several drug candidate compounds for the inhibition of binding between (SARS-CoV-2 S-protein RBD) Spike and ACE-2 (host cell binding target) proteins. We then compare the computationally-predicted results against the ones experimentally-observed in the laboratory. The relatively high agreement between computationally-predicted and experimentally-observed binding interactions supports the potential of our framework as the next-generation pipeline for prediction models in drug repurposing.

SARS-CoV-2 Spike Protein Extrapolation for COVID Diagnosis and Vaccine Development.

Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) led to coronavirus disease 2019 (COVID-19) pandemic affecting nearly 71.2 million humans in more than 191 countries, with more than 1.6 million mortalities as of 12 December, 2020. The spike glycoprotein (S-protein), anchored onto the virus envelope, is the trimer of S-protein comprised of S1 and S2 domains which interacts with host cell receptors and facilitates virus-cell membrane fusion. The S1 domain comprises of a receptor binding domain (RBD) possessing an N-terminal domain and two subdomains (SD1 and SD2). Certain regions of S-protein of SARS-CoV-2 such as S2 domain and fragment of the RBD remain conserved despite the high selection pressure. These conserved regions of the S-protein are extrapolated as the potential target for developing molecular diagnostic techniques. Further, the S-protein acts as an antigenic target for different serological assay platforms for the diagnosis of COVID-19. Virus-specific IgM and IgG antibodies can be used to detect viral proteins in ELISA and lateral flow immunoassays. The S-protein of SARS-CoV-2 has very high sequence similarity to SARS-CoV-1, and the monoclonal antibodies (mAbs) against SARS-CoV-1 cross-react with S-protein of SARS-CoV-2 and neutralize its activity. Furthermore, in vitro studies have demonstrated that polyclonal antibodies targeted against the RBD of S-protein of SARS-CoV-1 can neutralize SARS-CoV-2 thus inhibiting its infectivity in permissive cell lines. Research on coronaviral S-proteins paves the way for the development of vaccines that may prevent SARS-CoV-2 infection and alleviate the current global coronavirus pandemic. However, specific neutralizing mAbs against SARS-CoV-2 are in clinical development. Therefore, neutralizing antibodies targeting SARS-CoV-2 S-protein are promising specific antiviral therapeutics for pre-and post-exposure prophylaxis and treatment of SARS-CoV-2 infection. We hereby review the approaches taken by researchers across the world to use spike gene and S-glycoprotein for the development of effective diagnostics, vaccines and therapeutics against SARA-CoV-2 infection the COVID-19 pandemic.

Responses to COVID-19 in South Asian Association for Regional Cooperation (SAARC) countries in 2020, a data analysis during a world of crises.

Coronavirus disease (COVID-19) caused by SARS-CoV-2 was notified from Wuhan city, Hubei province, China in the mid of December 2019. The disease is showing dynamic change in the pattern of confirmed cases and death toll in these low and middle-income countries (LMICs). In this study, exponential growth (EG) method was used to calculate the real-time reproductive number (Rt) for initial and later stage of epidemic in South Asian Association for Regional Cooperation (SAARC) member countries (April 2020 - December 2020). Time dependent (TD) method was used to calculate the weekly real -time reproduction number (Rt). We also presented the observations on COVID-19 epidemiology in relation with the health expenditure, poverty, BCG vaccination, literacy population density and Rt for understanding the current scenario, trends, and expected outcome of the disease in SAARC countries. A significant positive correlation was noticed between COVID-19 deaths and health expenditure (% GDP) (r = 0.58, P < 0.05). The other factors such as population density/sq km, literacy %, adult population %, and poverty % were not significantly correlated with number of COVID-19 cases and deaths. Among SAARC countries, the highest Rt was observed in India (Rt = 2.10; 95% CI 2.04-2.17) followed by Bangladesh (Rt = 1.62; 95% CI 1.59-1.64) in initial state of epidemic. A continuous monitoring is necessitated in all countries looking at the medical facilities, available infrastructure and healthcare manpower, constraints which may appear with increased number of critically ill patients if the situation persists longer.

Epidemiology and pathobiology of SARS-CoV-2 (COVID-19) in comparison with SARS, MERS: An updated overview of current knowledge and future perspectives.

Severe Acute Respiratory Syndrome-Coronavirus-2 (SARS-CoV-2) is the causative etiology of 'Corona Virus Disease-2019' (COVID-19); formerly referred as 'novel-Coronavirus-2019'. It was originated in Wuhan city, Hubei province, China in early December 2019. The World Health Organization (WHO) declared it as 'Public Health Emergency of International Concern' due to their rapid transmission and causing public and health-care-related casualties worldwide. This review provides an updated overview of COVID-19 (SARS-CoV-2), in comparison with the etiologies of the same group viz. SARS and MERS and also its future perspectives for planning appropriate strategies for prevention, control and treatment modalities to avert similar catastrophe in near future.

CRISPR-Cas System: An Approach With Potentials for COVID-19 Diagnosis and Therapeutics.

COVID-19, the human coronavirus disease caused by SARS-CoV-2, was reported for the first time in Wuhan, China in late 2019. COVID-19 has no preventive vaccine or proven standard pharmacological treatment, and consequently, the outbreak swiftly became a pandemic affecting more than 215 countries around the world. For the diagnosis of COVID-19, the only reliable diagnostics is a qPCR assay. Among other diagnostic tools, the CRISPR-Cas system is being investigated for rapid and specific diagnosis of COVID-19. The CRISPR-Cas-based methods diagnose the SARS-CoV-2 infections within an hour. Apart from its diagnostic ability, CRISPR-Cas system is also being assessed for antiviral therapy development; however, till date, no CRISPR-based therapy has been approved for human use. The Prophylactic Antiviral CRISPR in huMAN cells (PAC-MAN), which is Cas 13 based strategy, has been developed against coronavirus. Although this strategy has the potential to be developed as a therapeutic modality, it may face significant challenges for approval in human clinical trials. This review is focused on describing potential use and challenges of CRISPR-Cas based approaches for the development of rapid and accurate diagnostic technique and/or a possible therapeutic alternative for combating COVID-19. The assessment of potential risks associated with use of CRISPR will be important for future clinical advancements.

BCG vaccine: a hope to control COVID-19 pandemic amid crisis.

COVID-19 caused by the virus SARS-CoV-2 has gripped essentially all countries in the world, and has infected millions and killed hundreds of thousands of people. Several innovative approaches are in development to restrain the spread of SARS-CoV-2. In particular, BCG, a vaccine against tuberculosis (TB), is being considered as an alternative therapeutic modality. BCG vaccine is known to induce both humoral and adaptive immunities, thereby activating both nonspecific and cross-reactive immune responses in the host, which combined could effectively resist other pathogens including SARS-CoV-2. Notably, some studies have revealed that SARS-CoV-2 infectivity, case positivity, and mortality rate have been higher in countries that have not adopted BCG vaccination than in countries that have done so. This review presents an overview of the concepts underlying BCG vaccination and its nonspecific immuological effects and protection, resulting in 'trained immunity' and potential utility for resisting COVID-19.

Coronavirus Disease Pandemic (COVID-19): Challenges and a Global Perspective.

The technology-driven world of the 21st century is currently confronted with a major threat to humankind, represented by the coronavirus disease (COVID-19) pandemic, caused by the severe acute respiratory syndrome, coronavirus-2 (SARS-CoV-2). As of now, COVID-19 has affected more than 6 million confirmed cases and took 0.39 million human lives. SARS-CoV-2 spreads much faster than its two ancestors, SARS-CoV and Middle East respiratory syndrome-CoV (MERS-CoV), but has low fatality rates. Our analyses speculate that the efficient replication and transmission of SARS-CoV-2 might be due to the high-density basic amino acid residues, preferably positioned in close proximity at both the furin-like cleavage sites (S1/S2 and S2') within the spike protein. Given the high genomic similarities of SARS-CoV-2 to bat SARS-like CoVs, it is likely that bats serve as a reservoir host for its progenitor. Women and children are less susceptible to SARS-CoV-2 infection, while the elderly and people with comorbidities are more prone to serious clinical outcomes, which may be associated with acute respiratory distress syndrome (ARDS) and cytokine storm. The cohesive approach amongst researchers across the globe has delivered high-end viral diagnostics. However, home-based point-of-care diagnostics are still under development, which may prove transformative in current COVID-19 pandemic containment. Similarly, vaccines and therapeutics against COVID-19 are currently in the pipeline for clinical trials. In this review, we discuss the noteworthy advancements, focusing on the etiological viral agent, comparative genomic analysis, population susceptibility, disease epidemiology and diagnosis, animal reservoirs, laboratory animal models, disease transmission, therapeutics, vaccine challenges, and disease mitigation measures.

Spatiotemporal dynamics in demography-sensitive disease transmission: COVID-19 spread in NY as a case study (preprint)

The rapid transmission of the highly contagious novel coronavirus has been represented through several data-guided approaches across targeted geographies, in an attempt to understand when the pandemic will be under control and imposed lockdown measures can be relaxed. However, these epidemiological models predominantly based on training data employing number of cases and fatalities are limited in that they do not account for the spatiotemporal population dynamics that principally contributes to the disease spread. Here, a stochastic cellular automata enabled predictive model is presented that is able to accurate describe the effect of demography-dependent population dynamics on disease transmission. Using the spread of coronavirus in the state of New York as a case study, results from the computational framework remarkably agree with the actual count for infected cases and deaths as reported across organizations. The predictions suggest that an extended lockdown in some form, for up to 180 days, can significantly reduce the risk of a second wave of the outbreak. In addition, increased availability of medical testing is able to reduce the number of infected patients, even when less stringent social distancing guidelines and imposed. Equipping this stochastic approach with demographic factors such as age ratio, pre-existing health conditions, robustifies the model to predict the transmittivity of future outbreaks before they transform into an epidemic.